The package xxl.core.math.statistics.nonparametric provides several methods based on nonparametric statistics. This includes especially complex estimators that are computed online over a data stream and aggregate the previously seen data. They are based on kernels or the empirical cumulative distribution function. Our generic framework also allows the integration of user-defined functionality. Concerning kernel estimation, different kernels are provided besides modellings of one- or multidimensional kernels and kernel based density estimators. Additionally, we provide some functionality dealing with histograms.